Towards an integrated approach to natural hazards risk assessment using GIS: with reference to bushfires.

This paper develops a GIS-based integrated approach to risk assessment in natural hazards, with reference to bushfires. The challenges for undertaking this approach have three components: data integration, risk assessment tasks, and risk decision-making. First, data integration in GIS is a fundamental step for subsequent risk assessment tasks and risk decision-making. A series of spatial data integration issues within GIS such as geographical scales and data models are addressed. Particularly, the integration of both physical environmental data and socioeconomic data is examined with an example linking remotely sensed data and areal census data in GIS. Second, specific risk assessment tasks, such as hazard behavior simulation and vulnerability assessment, should be undertaken in order to understand complex hazard risks and provide support for risk decision-making. For risk assessment tasks involving heterogeneous data sources, the selection of spatial analysis units is important. Third, risk decision-making concerns spatial preferences and/or patterns, and a multicriteria evaluation (MCE)-GIS typology for risk decision-making is presented that incorporates three perspectives: spatial data types, data models, and methods development. Both conventional MCE methods and artificial intelligence-based methods with GIS are identified to facilitate spatial risk decision-making in a rational and interpretable way. Finally, the paper concludes that the integrated approach can be used to assist risk management of natural hazards, in theory and in practice.